COMPUTER VISION

Edge Computer Vision Audience Analytics

All projects
Diagram of a digital sign with a camera counting viewers and their dwell times, next to an anonymous audience metrics panel

A modular analytics platform for retail and digital signage, deployed as software onto a client's existing kiosks and display hardware. It turns standard signage screens into data-driven touchpoints by anonymously measuring viewer engagement, traffic flow and impressions in real time.

Architecture and capabilities

  • Edge processing: computer vision runs locally on the display hardware, analyzing RTSP or USB video feeds to count people, track dwell time and gauge engagement without depending on cloud processing.
  • Offline-first resilience: local caching and automatic recovery queues keep metrics logging through network drops and full internet outages.
  • No new hardware: the platform installs as a software package that uses the kiosks' existing cameras and processors, with no physical installation or hardware changes.
  • CMS integration: kept separate from the core display logic, structured metrics are sent to digital signage CMS layouts, custom HTML widgets and admin dashboards over REST and MQTT.

Client impact

  • Privacy-first measurement: reports impressions, peak traffic windows and average dwell time without capturing or storing personally identifiable information or facial biometrics.
  • Measurable signage ROI: replaces guesswork with data on ad performance, audience engagement and screen occupancy.
  • Low-risk rollout: avoids hardware purchasing and logistics, so deployments can be pushed remotely across multi-site kiosk networks.

Technologies & tools

  • Edge computer vision
  • RTSP / USB video
  • Offline-first caching
  • MQTT
  • REST APIs
  • CMS integration
Antigua coastline and harbor at dusk
Have a similar problem?

Let’s solve yours.

Based in Antigua & Barbuda · Working locally and remotely

Start a similar project